Many studies of the human microbiome underemphasize the complexity of strain-level genetic diversity, partly due to computational challenges. This is an important problem because subtle genomic alterations can sharply impact microbial behavior, e.g. antibiotic resistance. Strain-level investigations are needed to understand how genomes change over time, and also to accurately characterize how microbial communities assemble, respond to perturbations, and vary among individuals. Our work focuses on the infant gut microbiome to address these fundamental biologic questions and to identify connections between infant health and early patterns of colonization. The objective of this project is to use strain-resolved metagenomic analyses to monitor microbial colonization in the infant gut during the first three years of life. We will test the hypothesis that early configurations of the infant microbiome can negatively impact maturation of the gut microbiome later in childhood. If this hypothesis is true, then manipulation of the microbiome in at-risk individuals, particularly premature infants, may provide opportunities to improve health outcomes. Our proposed work will characterize the population structure of microbial communities that develop during colonization of the infant gut and examine the roles of strain persistence, strain immigration, in situ genome diversification, and mobile genetic elements. To understand major temporal changes in strain or species abundance, we will utilize a novel method to infer microbial growth rates as well as community wide gene expression. We will conduct strain- level analyses of fecal samples from 100 newborn infants and their mothers during the first three years of life. We will include 40 preterm infants with no major medical problems, half born via caesarean section; 40 preterm infants that develop either necrotizing enterocolitis (NEC) or late-onset sepsis (LOS), half born via caesarean section; and 20 full term infants, half born via caesarean section. Deep sequencing of microbial DNA will enable genome reconstruction from coexisting bacterial, archaeal (if present), phage, and plasmid populations. This will allow us to track species membership, community structure, metabolic potential, and population-level genetic heterogeneity. We will use these data to test the hypothesis that some early-establishing strains persist beyond the initial colonization period (Aim 1); to test the hypothesis that stable gut microbial communities possess higher strain-level diversity than unstable gut microbial communities (Aim 2); and to test the hypothesis that clinical variables in the newborn period impact patterns of strain acquisition in the first three years of life (Aim 3). Improved understanding of community assembly and diversification in the infant gut could translate to improved outcomes by uncovering strategies for disease prevention and treatment. This research will also reveal universal principles of microbial community dynamics that will have implications for other aspects of human health and science.